Dead process

[CoffeeWithBurger]

Describe the bug

The scheme shows that the process branches into 2 directions after the start:

1. Some kind of load emulation (waiting 1 minute).
2. Timer, with a configured one-time call in a minute. But the timer is then restarted using the return (arrow).

If you start this process and interrupt the execution of the service after 5-10 seconds (before the user-task), then the process will not continue its execution, that is

1. The task for the user-task will never be created
2. The timer will turn off and never start again.

Because of these problems, the process becomes completely inoperable. Now the process can be restored only if it is in the user-task.

At the same time, all information about the process is stored in postgresql and it is expected that from there it will be able to recover independently.

P.S. Where can I ask questions about the implementation of any approach? The problem was found when I was trying to implement a mutable TTL of the process: depending on where the process is located (at the beginning, in the middle or almost at the end), its lifetime should change. P.S.S. I was also interested in how to put several 'Start messages' in one scheme so that it could be start from anywhere.

Expected behavior

The process is restored independently from the postgresql database:

1. The Service-task with simulated operation will start again and reach the user-task.
2. The timer will restart and the work will continue.

Actual behavior

The process does not recover independently from the postgresql database and becomes inoperable.

How to Reproduce?

1. Create a diagram as in the screenshot.
2. Start the process.
3. Restart the service 5 seconds after the start.

Output of uname -a or ver

No response

Output of java -version

17

GraalVM version (if different from Java)

No response

Kogito version or git rev (or at least Quarkus version if you are using Kogito via Quarkus platform BOM)

1.23.0.Final

Build tool (ie. output of mvnw --version or gradlew --version)

Maven 3.8.5

Additional information

No response

[CoffeeWithBurger]

The easiest way to reproduce the problem

[CoffeeWithBurger]

In the management console, you can see that the process with the error "java.lang.IllegalArgumentException - null source"

This error does not always occur: about once out of 5 Upgrading the version to 1.27 did not help - the error remained.